DOCUMENT
JSLHR_57_6_2261Supp_studies.pdf (187.93 kB)
DOCUMENT
JSLHR_57_6_2261Supp_table_a.pdf (207.84 kB)
DOCUMENT
JSLHR_57_6_2261Supp_table_b.pdf (131.25 kB)
DOCUMENT
JSLHR_57_6_2261Supp_table_c.pdf (127.99 kB)
DOCUMENT
JSLHR_57_6_2261Supp_table_d.pdf (125.02 kB)
1/0
Evaluation of Three Proposed Markers for Language Impairment in English (Pawłowska, 2014)
dataset
posted on 2014-12-01, 00:00 authored by Monika PawłowskaPurpose The goal of the study was to determine to what extent 3 proposed markers of language impairment (LI) in English (verb tense, nonword repetition, and sentence repetition) accurately distinguish affected and unaffected English-speaking individuals.
Method Electronic databases were searched for diagnostic accuracy studies involving the 3 markers. Quality of relevant studies was described. Numbers of true and false positives and negatives were extracted and used to calculate likelihood ratios (LRs).
Results Thirteen studies met the selection criteria. The majority were based on clinically ascertained samples. Pooled LRs and 95% confidence intervals (CIs) for tense (LR+) and sentence repetition (LR+ and LR?) were suggestive of presence (LR+) or absence (LR?) of LI. Wide CIs around the value of inconsistency I2 index reduced reliability of pooled values for sentence repetition. High between-study heterogeneity precluded pooling of LR values for tense (LR?) and nonword repetition (LR+ and LR?).
Conclusion The limited evidence available suggests that the proposed markers may be at best suggestive of LI in English. Future research may refine existing marker tasks to increase their accuracy and test the most promising tasks in unselected samples of participants with and without LI.
